POTENTIAL IMPACTS OF EXPLAINABLE ARTIFICIAL INTELLIGENCE (XAI) ON WATER RESOURCES MANAGEMENT: A SYSTEMATIC LITERATURE REVIEW
Description
Effective and sustainable water resource management is fundamental for human and environmental development. In this context, Artificial Intelligence (AI) emerges as a promising tool to enhance understanding, prediction, and decision-making in water management. However, the opacity of AI models, especially deep neural networks, raises concerns about the accountability and reliability of automated decisions. Explainable Artificial Intelligence (XAI) presents itself as a thriving approach to making AI models more transparent and understandable, particularly in critical systems such as water resource management. This study proposes a systematic literature review to analyze the state-of-the-art application of XAI in water resource management, identifying benefits, challenges, and knowledge gaps. The objectives include understanding the application of XAI in water management, synthesizing current knowledge, and discussing benefits and limitations. To achieve this, thematic filters were developed, followed by careful reading and categorization according to techniques, algorithms, and areas of application. A total of 314 studies published between 2021 and April 2025 were analyzed, with a predominance of the SHAP technique (in 247 studies) and the XGBoost and Random Forest algorithms. The most frequent themes were water quality, aquifer management, and extreme events. The main reported benefits include increased transparency, reduced subjectivity, greater institutional acceptance, and operational optimization. Among the challenges, the most prominent are methodological concentration, the lack of standardized evaluation metrics, and the need for greater interdisciplinary integration. This study reinforces the potential of XAI as a technical and ethical support for data-driven water governance, contributing to more informed, participatory, and resilient decision-making.
Files
Institutions
- Universidade Estadual de Campinas